Skip to content

reference_modifiers.md

Latest commit

 

History

History
162 lines (123 loc) · 3.66 KB

reference_modifiers.md

File metadata and controls

162 lines (123 loc) · 3.66 KB

reference_modifiers.md

Regular Expression Reference

PCRE compared to other flavors

MODIFIERS

g - global match

echo "abaca" | perl -ne 'while(/.?a/g){print "$&-";}'  # output: a-ba-ca-
```perl
# notice how the g modifier does not have an influence if not setted inside a loop
echo "abaca" | perl -ne 'if(/.?a/g){print "$&\n";}'    # output: a

i - ignore case
makes the pattern case insensitive

echo "A" | perl -ne 'print if(/a/i);'  # output: A

s - . matches also \n

echo "" | perl -ne 'print "yes\n" if/./s;'   # output: yes
echo "" | perl -ne 'print "no\n" unless/./;' # output: no

m - multiline match
if used ^ and $ also match begin and end of a line
does not influences \A and \Z
in perl overwrites the $* var

o - compile only once
Makes the regex execution faster.
Do not use this operator if the RegEx contain a variable which is changing when using the RegEx a second time.
If your RegEx doesn't contain variables, the modifier o is automatically on.

notice how in this example the $a variable is not changing because of the o modifier

echo "1234
1234
1234
1234"  | perl -ne '$a=$.; s/$a/X/o; print;'
# output: X234
# output: X234
# output: X234
# output: X234

instead of changing with every interaction in absence of the o modifier

echo "1234
1234
1234
1234"  | perl -ne '$a=$.; s/$a/X/; print;'
# output: X234
# output: 1X34
# output: 12X4
# output: 123X

x - extended free spaces and comments mode

#!/usr/bin/perl -w
use strict;

undef $/;
$_=<>;

s/^.+<table/<html><body><table/s;


s/<tr>[^<]+?
    
    <td>(.+?)<\/td>[^<]+?
    
    <td>(.+?)<\/td>[^<]+?

    <td>(.+?)<\/td>[^<]+?

  <\/tr>
/
    <tr>
    <td><br><h3 style='color:blue;'>  $1 <\/h3> <br><\/td>
    <td><br><h3 style='color:red;'>   $2 <\/h3><br><\/td>
    <td><br><h3 style='color:green;'> $3 <\/h3><br><\/td>
    <\/tr>
/sxg;

print $_;

e - execute
executes the content, even multiple times

echo "time" | perl -pe 's/time/localtime/e;'            # output: Sun Nov 29 17:17:22 2015
echo "A" | perl -pe '$var = "ord($1);"; s/(A)/$var/ee;' # output: 65

c in gc - keeps the current position during repeated matching

# notice how in this case using the gc modifier after the first regex the pos() is setted to 4 and
# how the second regex only outputs one "a" (the last one in the string)
echo "mamma" |
perl -ne 'while(/m/gc) {print $& . pos() . "-";} print "after: " . pos() . "\n"; while(/a/g) {print $&;}'
# output: m1-m3-m4-after: 4
# output: a
# in this case without the gc modifier the pos() after the first regex is not setted
# and the second regex starts matching from the beginning of the string finding a two times
echo "mamma" |
perl -ne 'while(/m/g) {print $& . pos() . "-";} print "after: " . pos() . "\n"; while(/a/g) {print $&;}'
# output: m1-m3-m4-after: 
# output: aa

in other words using the gc modifiers the pos() is not resetted after a failure match and the second regex starts when the first one failed

\G - start of match / end of previous match

In a similar way as the gc modifier you can use \G inside a regex:

echo "the-dog" | perl -ne 'if(/the-/g){if(/\Gdog/){print "following\n";} else {print "not following";}}'
# output: following
echo "the-dog" | perl -ne 'if(/the/g){if(/\Gdog/){print "following\n";} else {print "not following";}}'
# output: not following

In the example above \G forces the second regex to start matching just at the end of the first regex.
g modifier necessary for this to work.